Logo IMG
HOME > PAST ISSUE > Article Detail


Imitation of Life

Can a computer program reproduce everything that happens inside a living cell?

Brian Hayes

On Growth and Form

The WholeCell model is not greatly concerned with details of spatial organization. The metabolic module treats the cell as if it were a well-stirred reactor vessel, where all molecules have the same chances of interacting, regardless of their location. Transcription and replication enzymes occupy specific positions along the bacterial chromosome, but the coordinates are one-dimensional, measured with respect to the linear genetic sequence; they do not define position in three-dimensional space.

Nevertheless, the simulation does include a state variable for cell geometry, which describes the bacterium’s shape and eventual fission. Curiously, the shape defined by the simulation is not in fact that of the biological cell. M. genitalium is usually described as having a flask or pear shape—a ball with a single asymmetrical appendage. Including this detail would complicate the model without revealing anything of biological significance, so the simulated cell is given a simpler geometry. It begins as a small sphere and elongates into a cylinder with hemispherical end caps. At the end of the life cycle, after the two copies of the genome have migrated to opposite poles of the cell, the middle of the cylinder begins to constrict and then pinches off to form two new cells.

The rules of cell growth are not hard to understand: As the volume of the cytoplasm increases, the enclosing membrane must grow in surface area by a commensurate amount. The mechanics of cell division are more mysterious, but the model nonetheless gives a tentative account. The key component is a protein called FtsZ, which forms a ring girdling the cell in the plane where the two daughter cells ultimately part company.

comments powered by Disqus


Subscribe to American Scientist